02.04.2008 – Cognitive Consistency, the Endowment Effect and the Preference Reversal Phenomenon

April 2, 2008

02 April 2008Lille
Share this:Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Google+

Louis Lévy-Garboua –
Université de Paris I and Paris School of Economics

Date and place: 02/04/2008 at 14:00
IÉSEG School of Management – 7 rue Solférino – 59000 LILLE
Lecture Room: E21

IÉSEG Seminar
jointly with LEM UMR8179

Paper and slides:
Download the paper (PDF – 497068 Ko)

Co-Author(s) of the underlying paper: Serge Blondel

Description: We exhibit three non-standard cases of preference reversal (PR) between choices and valuations of bets. PR patterns diverge whether asking prices or bidding prices are used to elicit values, a finding not well explained by contingent weighting theory. PR also occurs when a lottery is confronted with a sure gain, which is inconsistent with regret theory. Lastly, PR does not require small probabilities of winning for the riskier bet. The endowment effect arises as a surprising consequence of the WTA/WTP disparity when the probabilities of winning get close to certainty. Due to the presence of a small uncertainty regarding goods which one does not own and to a discontinuity of WTP close to certainty, people underestimate the value of goods not owned but do not overestimate the value of goods owned. We consider a new theory of choice under risk, cognitive consistency (CC) theory (Lévy-Garboua 1999), which can predict both standard and non-standard cases of PR, the WTA/WTP disparity and the endowment effect in the context of single choices with a single parameter added to risk aversion. We show that the BDM mechanism and the random lottery incentive system are incentive-compatible if subjects have preferences which satisfy CC theory. A new experiment is designed to test the predictions of CC versus other theories of PR and the endowment effect. The experimental results strongly support CC predictions, reject the regret theory explanation for PR and the loss aversion explanation for the endowment effect, and allow direct estimation of risk aversion and the additional parameter of CC theory with no restriction on the utility function.